The broad, long-term objective of this proposal is to understand the etiology of a newly described disease, large granular lymphocyte (LGL) leukemia. Preliminary results show: 1) a specific antibody response to HTLV-I/II proteins in 54% of LGL leukemia patients; 2) detection of HTLV- related gene sequences in DNA from some LGL leukemia patients; 3) transmission of infectious retrovirus from a splenic LGL leukemia cell line to a human osteosarcoma cell line by coculture experiments. The specific aim of this proposal is to characterize the retroviruses present in cells or cell lines of LGL leukemia patients. The first approach will utilize PCR techniques to amplify retroviral gene sequences. Amplified products will be sequenced to determine homology to HTLV-I/II and then cloned and used as probes to screen genomic and cDNA libraries. Purified inserts from clones of interest identified in termination. Genomic libraries will be constructed from DNA of leukemic LGL or from retrovirally infected cocultured cell lines. As an alternative approach, cDNA libraries will be constructed directly from virion RNA banded on continuous sucrose gradients. Initial experiments will utilize the retrovirally-infected cocultured cell line, LGL 5-HOS, for retroviral isolation. To make this approach more feasible, cell lines which stably produce high titer retrovirus will be established using coculture techniques. Evidence of retroviral infection as determined by reverse transcriptase assays will be confirmed by electron microscopy, PCR techniques, and monoclonal antibodies to retroviral proteins using flow cytometry and radioimmunoprecipitation. Retroviral proteins will be further characterized by performing SDS- polyacrylamide gel electrophoresis directly on sucrose-gradient banded particles. Since there is a 20% mortality rate within four years of diagnosis and there is no effective treatment for LGL leukemia, understanding the etiology of the disease is essential for designing better therapy.